Towards Efficient Video Detection Object Super-Resolution with Deep Fusion Network for Public Safety

Algorithm 1

Keyframe selection algorithm.

	Input: # input video
	Output: # output keyframe
(1)	defineas backbone of the object detection network
(2)	defineas neck of the object detection network
(3)	defineas prediction of the object detection network
(4)	defineas output of the object detection network
(5)
(6)	# Choose category, quantity, BBox, confidence
(7)	for i = 1,... n do #n is the number of video frames
(8)	if (Ci! = Ci-1or Ni! = Ni-1)
(9)
(10)	Storage (Fi)
(11)	end for
(12)	#k is the number of key video frames
(13)	Return;